To treat genetic diseases by therapeutic gene transfer, it is usually necessary to integrated the therapeutic gene into a chromosome of the host cell. However, this has led to clinical adverse events in patients receiving gene therapy for SCID-X1, in which integration of retroviral vectors activated cellular proto-oncogenes, leading to transformation of gene-corrected cells. Thus the gene therapy community has become intensely focused on the question of where gene transfer vectors integrate in the human genome. The FDA has even mandated that integration sites be analyzed as a step in monitoring for possible adverse events. The Bushman laboratory has established a unique collection of technologies for analyzing integration site populations based on DNA bar coding, pyrosequencing, and custom bioinformatic tools. Using these high throughput methods, populations of integration site sequences can be generated of up to 105 bases of sequence in a single one day run. In one published study, we generated and analyzed the placement of 40,000 unique sites of HIV DNA integration in the human genome. Here we propose to apply these methods to systematic analysis of patients from the French SCID-X1 trial. We have initiated massively parallel sequencing studies of longitudinal DNA samples from some of the French SCID-X1 patients, with the goal of understanding the evolution, ecology, and ultimate fate of transduced cells. So far, we have generated ~128,000 integration site sequence reads from 80 patient samples for a total of ~33,000,000 bases of DNA sequence. For the first time, we can begin to estimate the numbers of transduced cell clones contributing to the gene-corrected cell pool. We can ask how populations of gene corrected cells change over time. In preliminary studies, we find a troubling decline in clone diversity in our two best-studied patients, suggesting "clone burn out". We can ask whether insertional activation of genes involved in growth control leads to outgrowth of clones even in the absence of clinical adverse events. For those patients who suffered adverse events and were subsequently treated by chemotherapy, we will ask how the chemotherapeutic treatment affected the size, diversity and dynamics of the gene-corrected cell pools. We will also analyze adverse events in animal models and, longer term, integration sites generated in new clinical trials. We propose to work in close collaboration with the French SCID-X1 team to complete the following Specific Aims: Aim 1. Determine the total numbers of integration sites present in samples from SCID-X1 patients. Aim 2. Determine how the number and distribution of integration sites changes over time in SCID-X1 patients. Aim 3. Determine integration site locations and relationship to genotoxicity in preclinical models and new clinical trials. PUBLIC HEALTH RELEVANCE: Success has been achieved with human gene therapy for SCID-X1, but adverse events due to insertional activation of proto-oncogenes and leukemia have caused severe setbacks. Here we propose to apply massively parallel pyrosequencing to analyzing vector integration in samples from the historic SCID-X1 trial.